9:00 A.M. in Shanghai, Shenzhen, and here in Hong Kong, and you're watching The China Show. Good morning to all of you. We are counting down to the opening of markets in Greater China. Let's get your top stories.
Happy day across equity markets for now, let's see how long it lasts. Reports about planning to spend $750 billion on additional computing power.
Meanwhile, a profit falls short and compounds worries at the stock pushed lower — the lowest in a year.
And fears around stocks could spread to the currency and bond markets. We will watch the open in Jakarta at the top of the next hour.
We continue to watch what goes on when it comes to the AI trade. Some tailwinds on earnings from SpaceX and the like. Talk about what is happening with Ellis, the CEO says he's not too concerned about volatility, but down under, what we are seeing with the data center and how they had to delay a project in Melbourne because of community backlash, and that stock is plummeting. We are seeing AI wobbles.
A better story today, but it's cleaner and more consistent than the U.S. in terms of stock performance. We mentioned breaking news in Indonesia, a rate decision today and a change to the timing. We are just finding out they are moving forward the rate decision to 1:30 p.m. local, should be earlier by a couple of hours, and I am wondering whether that would mean they make the decision when the cash market is open. I think how that conversation has changed has become almost purely because of market mechanics and dynamics around the currency. We are looking at a rate hike not because the economy needs it, but the markets perhaps need it more.
A two-hour earlier decision now, could be while markets are open. There are more calls and the consensus is they have to do something to stabilize the rupiah, in fear if they are dealing with a double downgrade when it comes to the MSCI. It's about how financial markets are doing, so even we are predicting a 50 basis point hike today.
You only need to look at the last meeting and they didn't do anything. That goes to show that nothing much can change in the economy in a month's time. What's changed is this, and we can put it down to that specifically. Equity markets doing well, should Indonesian markets? The index was up 5% into yesterday and we did see a pop in the market. They haven't come out with a decision on why. One could hope it has to do with the fact of what's going on and not down to other things like logistics. The U.S. 10-year yield, the Canadian dollar today had a good session on the back of Alberta voting not to leave Canada.
We talked about LS and we heard from the CEO.
They basically had to cut their guidance and talk about causing a lot of these projects when it comes to data center buildout. It does show there are challenges in Australia, but something we also see globally: the pushback you see from the public when it comes to building the data centers, whether it's water impacts, environmental concerns. It could be a bigger issue not just in Australia. It's one thing to watch.
We heard from the Elice CEO as well.
For the first day it was very volatile, and interesting for me to see how volatile it was in the first and second day. But in the long term we are looking at explosive growth. I'm not worried about volatile issues in the first week.
The stock is bouncing up about 30%, and it's worth looking at Chinese markets because there are a lot of other things going on. We have the challenged markets, and it's difficult to pick out the major earnings that came through, and analysts are cutting price targets for the stock. I expect some pressure to show up. We have a bond auction as well, a seven-year, if I'm not mistaken, and we are getting closer to the weekend. It's the APEC finance ministers' forum.
The finance minister of China spoke yesterday as well. Let's bring in our equity strategist. Things looking better when it comes to the tech trade, focusing on XTC as well. There are still AI nerves as well.
Good morning. The AI trade has been bifurcating for a few weeks, where some of the data center stocks have definitely been lagging or falling, and markets being rightly concerned. And the AI trade has been changing its course to higher-quality names like NVIDIA, which have held up better than the semiconductor space. There's general nervousness around what the business models, what return on capital they will make. With a hyperscaler, should it be trading on the same multiple as a neo-cloud? And that's come to the forefront the last few weeks. There's more of a discount that needs to be applied in those names. That's something, it's probably healthy the market has this bifurcation of two different trade ideas in the complex.
Despite, or maybe amidst that, the U.S. equity outperformance is out in front again. How do we explain that?
It was really interesting to see in Germany talking about regulation, and one of the comments was: build out the AI industry first before you regulate it. In the U.S., the attitude is a lot of pro-business and pro-markets and pro-AI, even with Trump's administration that's been quick to reassure investors they want to win in AI first. Then we see results out of SpaceX where capex has gone up, capacity buildouts are higher, reconfirming the compute shortage and GPU prices higher. Given a relief rally to markets that the AI trade still has legs, it might bifurcate to other pockets, and ultimately the U.S. is leading the buildout in gigawatts around the world.
What will happen with the SpaceX and Tesla merger if that happens? Sunny, thank you. Our equity strategist. Talking about what traders and China are looking at, we have to navigate weakness when it comes to the EV space. Let's bring in our markets reporter with the numbers we got.
Soft on the net income side. The company continues to expand into different business lines, trucks and the big fuel cells they need, but the relentless pace of competition in domestic China and meeting international headwinds on trade. The sector has been under the gun recently. Despite its market-leading position, it is not resistant. We have a decent set of numbers in terms of new models, it just seems like the pace of innovation is threatening the space as a whole. They are spending so much in R&D, it showed up in net income numbers.
How are batteries reflected?
Let's break it down. First is the impact on the market as a whole. EVs are still part of China's tech indices. EVs are around 20% of the Hang Seng Tech. The weakness has showed up in the performance, especially versus China and Taiwan. In terms of macro, they are also concerned because EVs were what AI was maybe five years ago. You have all this competition and no margins and no profitability. It's starting to remind investors of what AI looks like today. We have so many Chinese players and the technology is amazing, but where's the profitability? That's perhaps the knock-on macro message today.
Thank you, Anthony. On the back of that, we will see how CATL fares, already falling yesterday. Bloomberg Intelligence had a great report talking about the unwind in the China AI trade. What they are saying is the slide in A-shares since July is more of a crowded AI trade unwinding, or is it about profitability and earnings outlooks? They say it's more of the former, pricing out the unwind of the crowded trade. We could see a bounce.
Could be a rebound. The debate whether it's purely sentiment, and have we gotten to a point where earnings are not as good as a year ago? Has the selloff been slightly too much? What chart is this? The derating across the space. This is something we will look at later on. We will have more on the outlook.
We cut the price for A-shares, and we will have an update out of Indonesia. We know why they did move the meeting forward by two hours — to accommodate the governor's schedule, why not? We are counting down to the open of markets. This is The China Show.
Welcome back, you are watching The China Show. We are looking at strength across China, equity markets called higher, and the 10-year yield is closer to 1.6%, you read that right. A 667 handle on dollar-China. There are a lot of tech stories to get into, and we have Tencent, Alibaba, and ByteDance. Reportedly ordered more than 3 million NVIDIA graphics cards, and the sticker price tops $35 billion. Let's bring in Minmin Low for more on the purchase plan.
This is a type of graphics card that's fairly advanced. Graphics cards are a staple for retail consumers. You put them in your computers and they run steadily. Different from the GPUs that are enterprise-grade, they have hardware that allows them to talk to tens of millions of other GPUs. You can put them in the server to build your AI data center. According to Reuters, some of these companies are buying 3 million of these cards, and The Information previously reported the Chinese government may allow companies to acquire them. The reason why the U.S. is allowing it is because these cards, the memory, bandwidth, and computing throughput doesn't exceed a certain threshold. NVIDIA has said, according to Reuters, they won't be able to fulfill all of the orders, perhaps just 5% to 10%, so the demand is outstripping supply.
China is creating a state-owned data group to advance its AI visions.
It's called the China National Data Group, and this is significant because of the visuals. You have the optics and the vice premier inaugurating the opening of the center, and it's broadcast across the media. The signal is: if you are a company, you will want to work with this company because it's part of a top-down strategy. He reported the government is looking at dedicating a network of data centers, and the mission of this company is to prevent what they are calling data silos, to create a data network to promote cross-industry data sharing.
There seems to be so much going on in the tech space. They are opening up software to train AI models. What do we know about that?
This is significant because DeepSeek is releasing the inference software. Imagine you have an AI model that runs on hardware accelerators. Between the hardware and the model is the inference software. They are releasing it as an inference software, so if you are a company using a DeepSeek model, you are renting it. The cloud computing companies can run DeepSeek to enjoy cost-cutting because they are releasing the technology, and this is crucial for China because they have bottlenecks when it comes to the accelerators. The challenge is how you squeeze more performance out of the same amount of chips, and this is the answer.
Thank you. We heard from the Citadel founder and CEO, he believes AI could be as transformative as the internet was in the 1990s. Speaking exclusively from a conference, he added there are still major unknowns around the economics of AI models and the value they can ultimately deliver.
What we don't know is how, as these models become more intelligent, how much are they able to demand extremely high pricing because they help to solve the problems that have vexed humanity for centuries.
Because people don't know the answer to that, often the conversation goes to maybe this is a bubble. Ray Dalio says it is, Howard Marks says he's on bubble watch. I wonder if it's notable that this is a bubble, and if you have particular thoughts, if it's a bubble, if it looks like it will burst in certain parts of the different AI layers?
The big picture is there are certainly bubbles at this point in time. Any radical transformation of our economy is generally intertwined with a bubble, all right? Remember in 1999, 2000, the metrics we used to value internet companies — let's be clear, some of them had fanciful business models. I just wish I'd kept my Amazon stock. The internet, which was the great bubble, could it be more of a part of daily life today than it is? I've told this story before, but I was updating my Xbox with my son a few years ago. He looks at me and he goes, 'Dad, you lived before the internet existed.' I said, 'I did, I was born after the dinosaurs disappeared.' But yes, there was life before the internet, and now that's like an incomprehensible thought to a 20-year-old. These bubbles — if we want to use that word — often coincide with giant technological transformations. The railroads, for example. What's left in the wake of those bubbles, if we use that word, is often game-changing. The companies on the other side are the market leaders for years to come.
The Citadel founder and CEO, exclusively speaking with Bloomberg from the Future of Global Markets Conference. As we go into the session today, one name in focus, not down as much as perhaps the initial guidance was pointing to. 400 bucks a pop. Looking at other things across markets, we are doing better roughly speaking. There we go. Lots more ahead. This is The China Show.
We are looking at what happens with southbound flows. HSTECH, the 4000 level is one thing to watch. Given the tailwinds we are seeing, futures still in positive territory, so we are seeing a bit of stabilization and upside when it comes to the AI trade. Technicals like the China have been oversold, are we due for some kind of bounce back is the key question. Day two of the finance ministers' meeting in Hong Kong, and there's a briefing this morning as well. We talked about stocks and watches, Montage Tech as well. We also look at sectors, from chips and data centers given what we've been seeing. Australia too.
This rethinking: do I want a data center in my neighborhood?
I think the grid bottlenecks are becoming more of an issue. There's plenty of funding going into the investments. You're getting public backlash when it comes to how they will be able to continue with the buildout.
Midterms are coming up, and that's featured in Australia as well, they had to scale that down. Just I think this is a good perspective on what is causing the brunt of the drawdown in mainland China. The benchmarks that did well because of hardware, and hardware has been taken down more than a notch or two, whether that's optical names or battery names. We know about EVs. Any index with the bigger question of that has done exceptionally badly. CATL very much in focus today. The opening bell is next. We are watching the open of these benchmarks. This is The China Show.
Good morning and welcome back. The session just a couple of seconds away. The weather is fantastic — no, it looks fantastic but it doesn't feel fantastic. It's hot and humid today, it feels like summer is back. Which is to say, the optics look good.
Under the hood we are watching key earnings. CATL could be the disappointment today. Deterioration in the industry is one thing to watch out for. We have to deal with the AI trade, and it seems like today the SpaceX earnings and Super Micro earnings are helping things recover after what was pretty ugly yesterday. Markets are doing this right now, CSI 300. Still 2% from the bear market, but we might avoid that threshold. We are watching China and STAR 50, up 1% just 20 seconds into the trade. Hong Kong doing better, HSTECH is up 1.5%. We will watch what happens to CATL, there are so many different challenges: domestically, given the EV market is slowing down; overseas, with a cloudy picture; and we've been talking about the revival of the consumption tax in the battery space. 5% to 6% loss.
We don't have a long history of trading in Hong Kong because the data was just last year, but for context, we're getting close to some thresholds. The worst day for H-shares was about -8%. The worst day for A-shares was 2024, -14%. Suffice it to say we are at the level where we have to look at superlatives. We are continuing, and we can't seem to find a floor just yet on shares of CATL. We are approaching the 10% threshold, and we are just down 8% on H-shares. Let's look at other earnings. We will revisit that in a moment. Chip and hardware stocks in focus, we talked, Minmin went through the China tech news. Alibaba and Tencent and DeepSeek and what it's looking to do as far as software. Unit three coming up, 4% after a warning around revenue guidance given what's been going on. Let's talk about CATL. The China index is down about 40%. I think we have a lot of stocks on this, about a quarter of the points decline is down to a single stock. Where this goes, that will tend to influence where the market goes. That's H-shares specifically because it's the ChiNext.
Our next guest got quite a bit of trimming down. It's great to have you. Your take on earnings from CATL?
Thanks for having me. It's kind of below the market expectation. The revenue is on track, the expectation would be the relying of a quarter-on-quarter. We should have a more so-called low value adds. We believe that could be the domestic energy storage market, which is quite competitive, very fierce across competition. Its average margin level. And to highlight that there's some change: starting from April, there's a cut in tax refund rates. It's not fully reflected in the second quarter because of the difference between the shipment and revenue completion, and now it's fully reflected in Q3. Throughout September, China starts to levy the 2% consumption tax on domestic purchases. All of this combined, because the market expected previously there was sharing between suppliers and carmakers because of the tax burden, but it's not happening yet. It may be fourth quarter next year. That's my take on the results.
What was your take on the mood of the earnings call? Was management worried? Are these known knowns? Was it transitory? Is it calcified around the company?
Don't see they have too much concerns given the policies we've been aware of since earlier this year, and I think the demand side, we still believe the next few years. But it's uncertain for the next year. I guess it comes from overseas markets. There's the overseas energy market given the concerns. Elsewhere, driving long-term growth rates. There could be near-term pressures on profitability, and the demand or volume is still intact. We see that in the growth rates the next few years.
You mentioned uncertainty over ES for next year domestically, why is that?
There've been concerns over inventory levels for the batteries because there's a gap. There's been also concerns about returns given it could be different to ours, which could have lower returns. But we do see some positives with prices going down. If it continues into next year, the cost will be lower. That would be partially improving returns.
Demand hasn't really changed, it's about how the industry is looking at profitability. I think a lot of your peers are talking about the 0.0 RMB per watt-hour, is that the new normal or can it recover?
In the fourth quarter, maybe next year, this could be somewhat normalized. The new consumption tax starts from September, it will have an impact on the fourth quarter. For next year there's another step up for the consumption tax from 2% to 4% starting from September next year. We do see companies trying to improve the overall revenue mix, and we see more from domestic, but it could be high-end product sales with original mix change. Non-Chinese markets could improve the unit net profit in the long term.
The stock, the H-shares are trading at quite a premium to the A. How do onshore and offshore look at this, and why is there the gap?
Onshore investors, sometimes they are more sensitive to incremental changes of market sentiment. When I talk about offshore investors, long-term investors tend to believe in a high-quality name. They don't worry that much about potential challenges, not too many suppliers could feel that cap. For long-term investors, they still have this.
That's where I wanted to end the conversation. Whether it is batteries or solar or wind or EVs, help explain the dynamic because I can't wrap my head around it. What seems to be bleeding edge and scale in China doesn't seem to be translating to proportionate profitability and market performance. Why does that keep happening? And it's happening now, it happened to BYD and a few more companies, and it's not happening to CATL. What's the disconnect between what seems to be a good stock but it's not a good performer?
I would say they have a unique market position. They have a dominant market share in China, the company is having superior efficiency. It allows them to have better ASP and cheaper costs. I also want to highlight that. Also aware of concerns over overcapacity, and we are seeing new battery projects subject to stricter controls. They can't just expand the capacity without reaching a certain level. That's where we see the battery industries, still like we see growth with less concerns about overcapacity.
Ethan, thank you. China equity research analyst. We are approaching 10% losses on the stock, and it is hopefully temporary, still bullish. We are watching nickel producers said to be weighing output cuts in Indonesia. We will have more on the pressure facing the industry next. This is The China Show.
In the U.S., so they are going for the fourth quarter, they were aiming to raise about $5 billion, about 20 billion in value. Everything seemed fine until the last few weeks. Investors are watching very closely these data center companies, looking at volatility and earnings and prospects. And even though Day One has a track record, people are nervous and just watching markets closely and trying to size the right window before they go ahead with the IPO. They are very much full on looking at options. They want to go ahead, but obviously they don't want to risk in the market environment when they have the company to go ahead to potentially go sideways.
It's also how large and the pricing, there are other things bankers can tweak. What conversations have you had in terms of what other options issuers are considering if they want to stick to the timing?
That's a key conversation they're having back and forth. That is key right now. The options could be: OK, we go for the original target on value and size, but given what's happening in the market, investors are going to demand a steeper discount most likely before they put money to work, and that leaves you with two options: potentially lowering the valuation, which for existing investors is good, but also potentially doing a smaller-sized IPO, raising maybe 3 billion as an option. The other option is to sit tight and just delay a little bit the timeline and look for the right window to strike. Those options are under consideration.
Seems like the U.S. is the alternative destination right now. Speaking of what Manuel had mentioned, LS is recovering, the second day of the debut. 30% gains right now, close to regaining the ground we lost yesterday, but look in Sydney, talking about close to 37% losses given the cut to guidance, and they've been talking about postponing or pausing data center projects in Australia as well, given the public backlash.
A very timely conversation with the global head of ECM there. Lots more ahead, including this on the times on your screens. This is The China Show.
Welcome back. There we go, the performance letter across the benchmark here kind of looks — you know the algorithm? 4101, 1.5% to the upside. We are doing well. How well we will tell you after this break. This is The China Show.
Welcome back to The China Show. A look at the CSI 300, and we are pretty much flat on the benchmark. Keep in mind we are on bear market watch for that gauge. Seems like we have seen a bit of this reversal of sorts when it comes to the tech side. That is helping. Onshore markets are looking like this. Looks like we are losing a bit of ground, maybe CATL related.
It is heavily weighted on the ChiNext. That will have some weather disturbance if there is a tropical depression that is within the vicinity. That is why we have some humid weather, because there is a typhoon approaching parts of Southeast Asia. That is sidecar. Buying to the upside in South Korea. Hang Seng Index doing its thing of trading in a tight range, which it has done very frequently of late. Just to also mention as we change the page to CATL, we are looking at soft bond flows today. H-shares and A, down. If it seems to be performance parity today, it is the mirror opposite when you look at the other two. One yesterday.
LS was recovering after that plunge yesterday on its debut. Next in Sydney is the one to watch out for. This is the data center company that had to cut its guidance when it came to building up the data center in Melbourne because of the community backlash it received. That is an ongoing issue, the grid bottlenecks when it comes to this infrastructure buildout. That is hitting them in a big way. Other than that, things are looking good. In Indonesia we are holding on to the 5% pump on the back of the measures we heard from officials, tightening ownership when it comes to these stocks. We see the JCI up 1.3%. Bank of Indonesia is the next to watch in the next few hours.
That one-way meeting is shorter by two hours because in our back they announced they were moving forward with that meeting. It is happening two hours earlier than scheduled. By two hours. As you can see here in terms of the expectations around this specific meeting, theoretical mental exercise: if we did not have the MSCI issue pop up in the last meeting and this one, I would imagine this one would look more as a consensus hold. It is split, and you have a small minority that thinks they may need to move by a bigger amount.
A jumbo hike when it comes to our own Bloomberg Economics, saying they need to do something to address the investability concerns around this market. The MSCI decision is not until June. And it comes to the pressure on the rupiah, could extend. This is why the central bank may need to step in in a bigger way. Seems that we stabilized a little bit and contained the fallout. Tell us more about the outlook when it comes to Asian stocks.
There is some bit of stability, but we should not forget the peak of $500 million of market cap lost. This is some semblance of a low, but the story is not over. The clients may begin again as we look at how markets are assessing things. MSCI's risk emanating from what passive funds store MSCI, and what MSCI plans to do is now becoming a contagion risk for other asset classes. Rupiah has gone below the 18,000 handle. That is emanating a response for a tighter policy by the Bank of Indonesia. MSCI, most of the economies we polled were on hold. Things have changed since MSCI downgraded and led to a bigger rout in Indonesian stocks. Some expect a 25 basis point hike today. Bloomberg economists expect a jumbo hike. It does not look like the start of our tightening cycle, but it looks removed from what has been done in the past, which is direct intervention, to using the policy rate to stem risk towards rupiah. That also becomes a bit of a vicious channel which will further weigh on stocks. MSCI's downgraded risk has led to a lot of outflows. I think $500 million have moved out of stocks at this time. Bonds are seeing a bit of outflows. Context of raising, it might stabilize rupiah a bit. Since June, a sword is hanging on the equity market. There might be more outflows. MSCI Indonesia has about nine stocks. Except banks, most of the other stocks there are not liquid. It is a throwing out the baby with the bathwater kind of thing. There are a lot of gains in even illiquid stocks. Indonesia is making efforts to fix things, but looks right now a bit like managing the narrative rather than fixing the free float issue.
It is not like — okay, fine, the MSCI update did catch a lot of people off guard, but it is not as if they were not having these conversations with the exchange to get these reforms in. I wonder, in your view, do you think we have enough time for those reforms to finally be pushed through and take place in Indonesia?
I don't think so. There was time. They did take some steps, but it did not lead to a resolution of the main problem, which was manipulation of the stock market, high concentration, low free float. There is a timeline until June. June is when they will take the position. India is an economy that is four or five times what ASEAN or Asian economies are.
That economy took two to three years. Even if cycles of everything, sure, and with time, this is not enough to affect the free float issue. The market has almost $500 billion of market cap that has been lost. It looks like a tall task. Let's see what the Indonesian policymakers do. It looks interesting. They are trying to involve everybody. It is not like taking the problem head on. They would need more time. There's no time to sell this amount of an opinion of the fund managers in such a short span of time.
You also have been writing about this AI trade and how people are starting to differentiate more about what sort of valuation you put on a hyperscaler versus neo cloud versus a data center company. There seems to be a bifurcation in the whole AI trade. What do people say about neo clouds right now? How are people looking at the space right now?
